During FY08, we focussed on three subprojects. (1) Hepatitis B Virus Capsid Assembly. We study the HBV capsid protein which presents two of the three clinically important antigens - core antigen (capsids) and e-antigen (unassembled protein) - of this major human pathogen. After first showing that capsid protein self-assembles from dimers into capsids of two different sizes, we obtained, in 1997, a cryo-EM density map in which much of the secondary structure was visible, including the 4-helix bundle that forms the dimerization motif. This was the first time that such detailed information had been achieved by cryo-EM. Our subsequent research helped delineate the path of the polypeptide chain. We went on to investigate the antigenic diversity of HBV by using cryo-EM to characterize the conformational epitopes of seven different monoclonal antibodies raised against capsids. In FY09, we followed three lines of investigation. (i) We completed a project reported in FY08 in which surface plasmon resonance was used to measure the binding affinities of a set of murine monoclonal antibodies commonly used to discriminate between core- and e-antigen, including several that we previously characterized by cryo-EM. Unexpectedly, most antibodies bind to both antigens with high affinity. The exceptions are antibody e6 which detects an epitope accessible only on dimers and occluded on capsids, and antibody 3120 which detects an epitope presented only on capsids because its epitope spans an inter-dimer interface. (ii) To investigate the immunodominant epitopes recognized by the human immune system in infected patients, we isolated antibodies from plasma obtained from a HBV-positive patient (provided by Drs T. Heller and J. Liang, NIDDK), prepared Fabs from them, and studied their decoration of purified recombinant capsids by cryo-electron microscopy and 3D-reconstruction. Contrary to expectation, these polyclonal Fabs did not form a continuous canopy over the capsid but instead bound preferentially to two discreet sites;one located on the spike-like protrusions (sites occupied by the so-called immunodominant loops), and one in the region between these protrusions. These data indicate that the human IgG repertoire actually detects a rather limited set of epitopes very similar to those that we previously observed with murine monoclonal antibodies. (iii) Following up on the "native" high resolution mass spectrometry experiments reported in FY08 in which the masses of both size variants of the capsid were determined to within 0.1%, we pursued a collaborative project to measure the mechanical properties of both capsids by nanoindentation methods carried out by atomic force microscopy. The two capsids have similar overall stability and elasticity (Young's modulus of 0.4 GPa), but differ in their subunit exchange rates, as determined by mass spectrometry of capsids incubated with isotopically labeled subunits. The directional dependence of capsid elasticity was predicted from coarse-grained molecular dynamics simulation and compared with the nano-indentation data. (2) Assembly and Maturation of Bacteriophage Capsids. Our interest in capsid assembly lies in the massive conformational changes that accompany their maturation. These transitions afford unique insights into allosteric regulation. We study maturation of several phages to exploit expedient aspects of each system. The tailed phages afford an excellent model for herpesvirus capsids (see (2) above), reflecting common evolutionary origins. In FY08, we focussed on thre projects. (i) We used a combination of scanning calorimetry and cryo-EM to investigate the encapsidation of phage DNA, which represents an extreme case of genome condensation. Phage HK97 is well suited to study this phenomenon in view of detailed knowledge of its capsid structure. We found that, as filled capsids are heated, their DNA is released at relatively low temperatures (40 to 50 degrees). Heating increases the internal pressure, causing the capsid to rupture, releasing the DNA. DNA packaging also induces a change in the capsid structure that is reflected both in an earlier onset of thermal denaturation than empty capsids and in subtle morphological differences. (We also detected a similar effect in herpesvirus capsids - see (2) above). We envisage that this transition in the capsid shell is transmitted to the portal, altering its interactions with the packaging enzyme and thus signaling that packaging is complete. This project, outlined in annual report FY08, was completed and published in FY09. (ii) The capsids of double-stranded RNA viruses serve as specialized compartments for the replication and transcription of the viral genomes. We investigate the structural basis of this remarkable phenomenon in the phage phi6 system, which has a tripartite genome. In FY08, we published a paper describing the location of the P2 polymerase in the interior of the viral procapsid, as determined by cryo-EM of wild type and mutant particles. P2 is substoichiometric, occupying only 3 - 10 (depending on the mutant) of 20 potential sites. In FY09, we extended these studies by cryo-electron tomography with particular focus on the question of whether sites occupied by P2 correlate with external sites occupied by P4, the packaging ATPase. The observed distributions of P2 and P4 occupancy indicate that both proteins are randomly distributed and therefore there is no direct coupling between the activities of these two viral enzymes that respectively conduct RNA packaging, and replication and transcription. (iii) For tailed phages and herpesviruses, a scaffolding protein is essential for correct capsid assembly . The scaffolding protein coassembles with the capsid protein to form the precursor procapsid and is then expelled during DNA packaging. Little is known about how scaffolding proteins perform this function. We study this phenomenon in the phage T7 system. Mass analysis of T7 procapsids indicates that the gp9 copy number averages 150 but can vary by as much as 30% from particle to particle. Examination of native T7 procapsids by cryoelectron tomography revealed a loosely ordered network of scaffolding filaments, with as many as ten filaments per particle. These filaments emanate from the portal vertex. The extended structure and positively charged C-terminus of gp9 likely play an important role in these interactions. The implications of these observations for T7 morphogenesis are under assessment. (3) Papillomaviruses and polyomaviruses. These viruses, which include human pathogens that induce cervical cancer, have the unique property of appropriating histone proteins from the host cell when assembling the viral minichromosome. In FY08, we started a project to analyze the structure of encapsidated SV40 minichromosomes by cryo-electron tomography. In a set of 100 reconstructed full virions extracted from several high-quality reconstructions, we find that the number of nucleosomes that they contain varies from 17 to 23 and that their spatial arrangement does not match the icosahedral symmetry of the capsid. Furthermore, analysis of extracted chromatin core densities suggests that overall chromatin structure is variable from particle to particle. Ongoing efforts are directed towards computationally modeling of the positions and orientations of nucleosomes within individual particle reconstructions.